Central Bank of Rift

Video game companies rarely hire economists. There are a few notable exceptions; specifically, EVE Online is known for keeping one on staff. But mostly these companies use crude (yet effective!) tools to manage their game economies.

• Strict capital controls or simply, a lack of inter-game trade. 
• Predictable inflation
• The gaming equivalent of selling government bonds: Gold sinks

The formula is simple. Players do things and earn in game cash. This is the primary source of inflation. Gold sinks exist to sop up some of this. The lack of inter-game trade keeps the whole mess from getting out of control. If you allow in game currencies to be converted to "real" money, or vice versa you run the risk that the whole house of cards will collapse on you. Diablo 3 is an example of the dangers of hyperinflation in a virtual economy. 

Hyperinflation is interesting, but easily solved. You can either:

1. Stop "printing money" by paying players less for the things they do (and watch the proletariat revolt) 
2. "Sell bonds" by putting in bigger sinks. (I'll trade you these new shiny pixels for that gold you're carrying)
3. Seize select accounts and declare the funds illegal. Ahem, I mean "against the terms of service". When governments do this, it makes headlines.

All this is fairly well understood. So when Rift decided to go free-to-play and more importantly, allowed a way of purchasing credits with in game currency out came the hyper-inflationary conspiracy theorists. Rift is doomed to hyperinflation because everyone will do the following:

1. Player 1 spends real money to get credits. 
2. Player 1 lists these credits (in the form of REX) for sale in game. 
3. Player 2 has some excess plat and buys the credits
4. Player 2 uses credits to buy items that could be sold to a vendor for even more in game currency. Magic inflation because now both players have lots of plat and they can double their plat indefinitely like this.

Questionable math aside, if the phenomenon exists all it does is set a floor on the value of REX. Furthermore, it ignores how people actually use REX. People without in game currency buy REX to sell to others who have an excess of in game currency. These people are unlikely to convert credits into even more platinum - they already have platinum. Instead they consume the credits, buying things from the shop. Also, the cash shop introduced a huge number of cash sinks and because of the REX conversion these are also platinum sinks. The net effect of all this is primarily deflationary.

Most importantly - I believe it is the intention of the game developer's for REX to be deflationary, at least in the short term. The developers will work to keep a balance between encouraging people to spend real dollars to buy REX and sell it to others and keeping REX remains affordable to the vast majority of players. Currently on my server (Faeblight) the price of REX has stabilized between 950 and 1050 and there is some speculation that the price will start to inch up from here. I think if REX prices spike too high, you will see creative moves by the developers to bring it back down. One likely move is to add in valuable in-game items that must be purchased with plat instead of credits.

In other words, what we are seeing in Rift is an impressive move to shift the in game economy. These changes both give the developers more levers to adjust and make it easier for players to see the effects of those changes on the economy through a solid connection to real dollars. For those out there who follow game economies and their parallels in the real world, Rift is a story worth following.

FINRA Data Harvester Revisited

I first talked about this program almost a year ago but I never posted it for others to look at and use even though the intention was to make my data tools as free and available as possible.

I have now set up a new Google Code project to host my financial data harvesters and the Regulation SHO harvester has been uploaded. Running it currently requires a few too many dependencies to be very user friendly, but if there is interest I will make a more user friendly version.

You can find it here:
http://code.google.com/p/finance-data-harvester/

Expect more such code + some data analysis soon.

Fight with Tools

I'm about to graduate and I needed to face the fact that I would no longer have access to some of the awesome tools that I get to use as a grad student.

Here is what I will no longer have access to that I use constantly:
CapIQ (+Excel Plugin!!)
Crystal Ball
Morningstar
Thompson One
S&P Backtester

Here's some stuff I have access to that I rarely use but will probably miss anyway:
Bloomberg Terminal
WRDS
Datastream

One of the goals of Sentinel Research is to provide analysis with cheap tools. I don't have millions of dollars under management, so I can't be spending capital on these kinds of things. So I've started looking into what I can find online.

FinViz.com is a decent free stock screener - it obviously doesn't have everything CapIQ does, but it has the basic screens I run most often.

I have found that Yahoo! Finance is close in format and info to Morningstar and to be completely honest with myself a lot of what I use Morningstar for does not require full access.

I am currently looking for good replacements for Crystal Ball and Backtester but I haven't found suitable candidates yet.

Beyond these, I actually plan on moving into more of a machine learning approach to stock research, so I plan on spending my time in R, SAS, and Python. My goal is to design some simple tools to evaluate stocks based on criteria that I believe to influence returns, then to test the results. My best ideas will be backed with real capital but I will keep track of my secondary ideas using a "fantasy portfolio" as well.